A goodness-of-fit test based on empirical processes is proposed as a model diagnostic check method for continuous time stochastic volatility models. More specifically, as the volatility is not observable, a marked empirical process is constructed from the representation in a state space model form associated to the discretized version of the underlying process. Distributions of these processes are approximated using bootstrap techniques. Some simulation results and an empirical application to an EURIBOR (Euro Interbank Offered Rate) data set are presented for illustration.